According to George Hripcsak, MD, author of the article and a professor of biomedical informatics at Columbia University in New York City, data in EHRs is inaccurate because information does not account for changes in patient conditions over time such as between night and day. In addition, patient symptoms may be poorly documented before death, according to a eWeek report. Dr. Hripcsak also told eWeek.com that the healthcare industry needs new models of healthcare processes to understand how to use its data properly.
Dr. Hripcsak and David J. Albers, co-author and associate research scientist of biomedical informatics at Columbia, recommend that researchers study EHRs as objects of interest to efficiently extract clinically relevant data currently locked in health records. In addition, more collaboration among researchers in informatics, computer science, statistics, physics and epidemiology are necessary to better comprehend EHR data, according to the article